Log file Viewer
#mongodb logs for Tuesday the 22nd of October, 2013
[00:45:01] <snkcld> for example, favorite_artists, such that when a new document is created, it would always have that field?
[00:55:24] <ruphos> I'm trying to migrate utf8 encoded text from the subdocument of one collection to its own collection with the perl MongoDB driver. It throws errors on the utf-specific characters ("Wide character in subroutine entry at /usr/local/lib/perl/5.14.2/MongoDB/Collection.pm line 168"). Is this just an issue with the perl driver, or is there special handling to be done with utf8?
[01:01:12] <vl4kn0> Is there a way to do 'binary search' instead of fulltext search? Meaning fulltext search except accepting all values ranging from 0 - 255 ?
[02:41:27] <cesarkawakami> vl4kn0: "fulltext" usually means taking into account some characteristics of natural language. Which will naturally not be present in binary data. What do you want to do?
[03:30:20] <raar> How do I set the default connection parameters of my mongo client? I want to change the default port to 27018 instead of 27017..
[03:38:40] <cheeser> there's no clear consensus on if the client uses that config file or not but if it doesn't (which probably isn't likely) then you're SOL on a permanent configuration. you'll just have to pass an option each time
[03:53:57] <jfisk87> the key is detectedApps, its not getting an error but it isnt updating the field
[03:58:45] <joannac> jfisk87: in your code, if you replace the call to update with a call to find, does it find the correct thing?
[11:16:21] <lqez> Thanks- Nomikos, Nodex- I created it for my MongoDB-based web applications. (dog food kkk)
[13:23:44] <singler> hello, I just noticed that there are not images in the docs. Is it a known problem?
[13:52:43] <banditron> diagrams aren't loading in the docs. e.g. http://docs.mongodb.org/manual/core/crud-introduction/
[13:53:15] <CupOfCocoa> Hey guys, I use pymongo as my driver and want to access the id of some documents in a cursor. specifically, I want to build a list of all the document ID's from the cursor to insert that list into a new document, so basically my document 2 contains a grouping of some documents 1 with some additional info. How would I go about this? A straightforward [p['id'] for p in cursor] leads to nothing
[14:10:58] <vl4kn0> Is there a way to use mongodb search facilities to search for any other than string values? I've got N integer values ranging from 0 to 255 and need to do fulltext search on them (the values are not actually any meaningfull strings)
[15:17:39] <vl4kn0> Nodex: still not what I'm looking for, I need to look for a vector for an occurence in a smaller vector, not any of the elements within the vector
[15:27:25] <guilhermekfe> how should I proceed in order to filter the documents in that format http://pastie.org/8421600
[15:32:03] <Nodex> vl4kn0 : perhaps you could explain it in better english than that, I don't understand complex words
[15:35:12] <guilhermekfe> Actually I'm using "User.find({ _id: user_id, 'projects.project_id': { $in : project_id } }, function(err, user) {..."
[15:35:30] <guilhermekfe> But I get this error, "[TypeError: Object 5265c8c65060a0180b000002 has no method 'map']"
[15:44:35] <guilhermekfe> But I think I know the issue... I changed the type of project_id in User Schema
[15:45:04] <guilhermekfe> I'm very new to no-sql concepts, and I have this doubt regarding one-to-many relations
[15:46:33] <guilhermekfe> My doubt is, should I use String for project_id, and then take out ref... or is this the best way?
[15:52:02] <salty-horse> hey. I'm pondering what's faster: whether I should perform two updates, one of which produces a duplicate key error, or instead do a find() followed by a single update. Is there a performance penalty because of the error generated on the server?
[15:53:25] <Derick> it will update when it matches the criteria, and if it doesn't find a doc matching the criteria, add one
[15:54:31] <salty-horse> derick: The first update mentioned is an upsert with a conditional: So it can either (1) succeed silently, (2) perform an insert, or (3) fail with dup key error due to the full conditional not matching any and insert failing for the duplicate key. that's not my question though :)
[15:55:23] <Derick> find + single update I think is better, but I'd say try it with production load and data :-)
[15:55:33] <salty-horse> I'm just wondering if the duplicate key error is tasking the server somehow. maybe it needs to return extra data to the client, and I'm better off just doing the find() first doing things right :)
[16:49:24] <fuzzi> my client (mongodb-native, jscript), dropped the connection, and now my collection is corrupted.
[17:28:09] <mboman> How can I query a subdocument? I have pefunctions.[list of dlls].[list of functions] and I want to query for a specific function but I don't know (beforehand) which DLL it is located in
[17:42:57] <mboman> jyee, not sure I understand the document. I've pasted an example document at http://pastebin.com/BnjDfqDs which is from db.metadata.find().limit(1). How can I query for "IsDebuggerPresent" in pefunctions.* without knowing that I should be looking in KERNEL32?
[17:51:07] <rh1n0> Anyone seen where a version upgrade causes 0 objects to be returned (using mongoid). saving/writing works fine. We just upgraded to 2.4.6 on ubuntu. tx
[17:56:15] <jyee> http://stackoverflow.com/questions/6790819/searching-for-value-of-any-field-in-mongodb-without-explicitly-naming-it
[17:57:36] <jyee> i was going to say that it's not possible, except for doing it app side, but those ^^ suggest with a schema change to use an array, you could do it.
[17:57:48] <retran> can do a map reduce where you shove all the data into a single field, and the ObjectId in a separate field
[17:58:21] <retran> then search on your single field, and get returned the ObjectId of the original object
[19:04:06] <squeakytoy> Need some suggestions. I have two collections; Users and Videos. Now a user should be able to "like" a video. Should I store that in the Video document? { likes: [ *user ids* ] } ? That will be a crazy long list for each Video Document.
[20:01:31] <joshua> Does this help http://docs.mongodb.org/manual/reference/operator/aggregation/geoNear/#pipe._S_geoNear
[20:04:29] <Frnk> but i think I could calculate something like: difference = radius - calculatedDistance and then query for: difference > 0 or somethink loike that
[20:30:00] <Frnk> db.test.aggregate([{ $geoNear: { near: [45, 11], distanceField: "distance" } }, {$project: { 'diff': { $subtract: ["$distance", "$radius"] } } }, { $match: { diff: { $gt: 0 } } }])
[21:19:51] <dxd828> Hey, something is trying to run "profile: -1" every few seconds as I can see in the logs. But the user keeps getting permission denied. What do i need to give that user so it can run this?
[21:41:42] <angasulino> I'm using the java driver to load an object from mongodb, it has a set_id(ObjectId) and ObjectId get_id(), but the ObjectId is not being loaded (I want to put the Id in some HTML so I can get it back when it's selected), I have searched but haven't come up with anything even related to my question, so I must be using the wrong terms, what am I doing wrong?
[21:57:15] <angasulino> cheeser, thank you! I'm learning ant, spring, mongodb and jslt all at the same time, sometimes I make silly mistakes :-)
[22:49:26] <LoneSoldier728> basically, I guess I want to name mongoose collections on the fly, I know they are specified in the model usually
[22:50:52] <LoneSoldier728> what makes more sense - basically these are the type of data I am going to store --- channel, week, year, songs - i was thinking of making a collection of channels, then within each channel have a collection of years, then in that a collection of weeks, and then a collection of songs within that
[22:52:53] <retran> then if you need them in one collection, do a map-reduce as needed, or "join" them in app
[22:53:17] <LoneSoldier728> well I would be querying everything that has a specific channel, week and year
[22:53:41] <LoneSoldier728> so i would query mongo with those specific fields and asking for it to spit back what it got for me
[22:59:16] <tripflex> https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCwQFjAA&url=http%3A%2F%2Fmongoosejs.com%2Fdocs%2Fpopulate.html&ei=CQNnUoOUEMP1kQfLjoDIBA&usg=AFQjCNHTC30cu9eYc298-HnXiM8E99-sNw&sig2=YWlGbOc__97Yw9c6bbjzyQ&bvm=bv.55123115,d.eW0
[23:22:43] <disorder> I have a list of tags, and I want to check if every tag has attached a document, if not I want to delete that tag from its own collection
[23:22:56] <disorder> collection.findOne({'metadata.user_id': req.user._id,'metadata.tags':tags[i]},function(err,item){
[23:22:56] <disorder> if ((item==null) && (!err)) tags_coll.remove({'user_id': req.user._id,'name':tags[i]},function(){})